Interactive test · five minutes
What colour is the letter A?
For some people letters have colours: A is red, B is blue, always the same, for as long as they can remember. This form of synaesthesia, in which letters have colours, turns up in roughly one or two people in a hundred. Give each letter the colour that seems to be its own. Every letter comes up three times, in a shuffled order; at the end you see how consistent you were and how your alphabet looks next to the 6,588 people with synaesthesia in an American study.
Tap the colour that feels like the letter's. Do not think too long: the first impression counts. If a letter has no colour for you, press "No colour".
The colours you pick stay in this tab. The page sends nothing anywhere.
Your alphabet
Your alphabet appears here once every letter has its colours.
What it means to see letters in colour
For people with grapheme–colour synaesthesia, each letter has a particular colour, which they see or simply "know", and which stays the same for years. The test here checks one thing: whether you give the same letter the same colour three times without knowing when it will come.
The test here uses the consistency measure of the Synesthesia Battery by David Eagleman and colleagues, published in 2007 and run ever since at synesthete.org. The score is the distance between the three colours picked for the same letter, averaged over the letters. Picks made completely at random give about 3. The Battery counts a score under 1 as "as consistent as a synaesthete"; a 2013 study proposed a cutoff of 1.43. All 6,588 people in the data passed the cutoff of 1, with a median score of 0.68.
The alphabet of the 6,588
Each row below is a letter, and each stripe is the colour one of the 6,588 gave it, arranged by colour family. For some letters one colour clearly leads: A is red for 53% of them, Y yellow for 44%, B blue for 40%. O and I are most often white, X and Z most often black. Others, such as J, K or U, have no clear favourite.
The fridge magnets
In 2015 Nathan Witthoft, Jonathan Winawer and David Eagleman searched these alphabets for the traces of a toy: a set of Fisher-Price magnetic letters, made in the United States from 1971 to 1990, in which the colours repeated in the order red, orange, yellow, green, blue, purple:
For 400 of the 6,588, at least 10 of the 26 letters are in the same colour family as in the set. If each person shuffled their own colours among the letters, that would happen less than once in a thousand. Of these 400, the 388 with a known year of birth were all born in 1967 or later, so they could have had the toy as children. Among those born from 1975 to 1979, 15% have the magnets' alphabet.
The authors draw a careful conclusion: in some people with synaesthesia, the colours were taken from whatever was around them when they learned their letters. The calculation here, rebuilt from their data and code, gives the same numbers: 400 people, and 329 of the 3,618 born in the years the toy was made.
What the page cannot tell
- It does not diagnose anything. A score under the cutoff only means you picked very consistent colours in one sitting. Researchers repeat the test after a while and use other checks too. A score above it does not mean you do not have synaesthesia either.
- The 6,588 are people in the United States who came to synesthete.org on their own and passed the test, so they are not a sample of the population. Their data hold one colour per letter.
- Screens show colours differently, and people with colour-vision deficiency (about one man in twelve of European descent) may see and pick colours differently. Each letter's colour family comes from the authors' colour grid, which has 11 colour names.
- The page guesses nothing about you from your colours, neither your age nor where you grew up. The toy in the study was American.
Method
- The data are those the authors published on Zenodo (GPL v2 licence): for each of the 6,588 people, the colour of every letter in red–green–blue values and, for 6,232 of them, the year of birth. The page publishes only totals and distributions, never anyone's alphabet.
- Your score follows the Battery: for each letter the distances between the three colours picked are added up (the sum of the absolute differences in red, green and blue, each from 0 to 1), then averaged over the letters. A letter given a colour once and "no colour" once gets a distance of 1 for that pair, what a random pick gives on average. Letters with no colour every time are left out of the average.
- The colour picker covers every colour a screen can show: hue and lightness on the surface, saturation on the slider, plus a grey band. Random taps on it would give an average score of 3.86.
- Your colour for a letter is the average of your three picks. Its family (red, blue, brown and so on) comes from the authors' grid, the same one used for the 6,588; checked on all 171,288 colours in the data, the page gives exactly the same families.
Sources
- Nathan Witthoft, Jonathan Winawer, David M. Eagleman, "Prevalence of Learned Grapheme-Color Pairings in a Large Online Sample of Synesthetes", PLoS ONE 10(3): e0118996 (2015); data and code on Zenodo and GitHub, GPL v2 licence.
- David M. Eagleman, Arielle D. Kagan, Stephanie S. Nelson, Deepak Sagaram, Anand K. Sarma, "A standardized test battery for the study of synesthesia", Journal of Neuroscience Methods 159 (2007).
- Nicolas Rothen, Anil K. Seth, Christoph Witzel, Jamie Ward, "Diagnosing synaesthesia with online colour pickers: Maximising sensitivity and specificity", Journal of Neuroscience Methods 215 (2013).
- Duncan A. Carmichael, Mark P. Down, Richard C. Shillcock, David M. Eagleman, Julia Simner, "Validating a standardised test battery for synesthesia: Does the Synesthesia Battery reliably detect synesthesia?", Consciousness and Cognition 33 (2015).